ASIAN MYRMECOLOGY Volume 10, e010006, 2018ISSN 1985-1944 | eISSN: 2462-2362 © Naoto Idogawa and Shigeto Dobata

Colony structure and life history of Lioponera daikoku(Formicidae: Dorylinae)

Naoto Idogawa* and Shigeto Dobata

Laboratory of Insect Ecology, Graduate School of Agriculture,Kyoto University, Kitashirakawa-oiwakecho, Sakyo-ku, Kyoto 606-8502, Japan.

*Corresponding author: idogawa.naoto.na@alumni.tsukuba.ac.jp

ABSTRACT. We report the colony structure and life history of the basal doryline ant Lioponera daikoku in Kyoto Prefecture, Japan. The following characteristics were observed in the field: (1) nests were always in the cavities of dead bamboo stems on the forest floor; (2) the colonies were small, averaging 10 workers with a single dealate queen; (3) each queen had four ovarioles (2 + 2), while workers had two (1 + 1); (4) alate sexuals with normal wings were produced in August, and participated in nuptial flights the next May; (5) brood production showed no indication of synchronization, and immatures at different stages coexisted in the same nests; and (6) the larvae and pupae of Monomorium ants were stored in the nests. Although many ant species nested in the bamboo stems at the study site, we suspect that L. daikoku is a specialist predator of Monomorium ants. We also describe the male of this species.

Keywords: Cerapachys, Ooceraea, army ants, raider ants, bamboo-dwelling ants

DOI: 10.20362/am.010006

INTRODUCTION

The subfamily Dorylinae Leach, 1815 is a group of monophyletic predatory ants distributed main-ly in tropical and subtropical regions (Brady et al. 2014, antmaps.org 2018). The subfamily encom-passes the so-called army ants and raider ants. Some army ant genera, including Aenictus, Eci-ton, and Dorylus, possess a peculiar set of traits, often known collectively as army ant syndrome, including a nomadic life history, extreme queen–worker dimorphism, and specialized mass preda-tion (Schneirla 1971; Gotwald 1995). Their social organizations have been studied extensively as a typical example of a superorganism (Camazine et al. 2001). Recently, the draft genome sequence of the clonal raider ant Ooceraea biroi Forel, 1907 (previously Cerapachys biroi in the subfam-ily Cerapachyinae) was determined (Oxley et al. 2014), making this subfamily an emerging model

system of integrative biology that covers studies ranging from molecular mechanisms to social or-ganization (Kronauer 2009). For a deeper understanding of the diver-sity of this subfamily, comparative studies are es-sential: detailed information on the species’ biol-ogy and their phylogenetic relationships should yield insights into the adaptive significance of their peculiar life history and the evolutionary history of their diversification. A recent taxonom-ic revision (Borowiec 2016) and large-scale mo-lecular phylogenetic studies (Brady et al. 2014; Borowiec 2017) have made this possible. The genus Lioponera was originally de-scribed by Mayr (1879), but was later downgrad-ed to a subgenus in the genus Cerapachys (Brown 1975). Recently, (Borowiec 2016) revived its tax-onomic status as a genus in the subfamily Doryli-nae based on molecular phylogeny (Brady et al. 2014). Lioponera is one of the phylogenetically

2 Naoto Idogawa & Shigeto Dobata

basal groups in the Dorylinae (Borowiec 2017), and all known species are specialist predators of ants (Borowiec 2016). This group is the most species-rich outside of true army ants (Borowiec 2016), which are distributed across the Old World. Here, we report the basic life history information of L. daikoku Terayama, 1996, which is one of the northernmost distributed of the Dorylinae species (Terayama 1996; antmaps.org 2018). We report on colony composition, colony size, nesting sites, seasonal production of sexuals, and prey items. We also provide the first description of the male of this species.

MATERIALS AND METHODS

The study site was a thicket consisting mostly of deciduous broad-leaved trees located in the northern suburbs of Kyoto, Japan (35°03’33” N, 135°47’01” E, alt. 90 m). We collected L. daikoku and other ant species nesting in the cavities of dead bamboo stems during field studies conduct-ed between 5 May and 27 November 2017. The numbers of each caste (dealate and alate queens, workers, and males) in the nests were counted in the laboratory immediately after sampling. Other species collected were identified using keys from Terayama et al. (2014) to clarify the myrmeco-fauna of the study site, because several studies of Lioponera ants suggest that species in this genus are specialist predators of other ants (Wilson 1958; Hölldobler 1982). Between 8 August and 27 No-vember, the diameters of bamboo nest stems were measured using digital Vernier calipers (accuracy 0.1 mm, DT-150; Niigata Seiki) to investigate nest site preference for all ant species collected. We photographed each adult under a digital microscope (VHX-900; Keyence). As in-dicators of body size, the widths of the head, tho-rax, and abdomen were measured using ImageJ software (Open Source: http://imagej.nih.gov/ij/). Some queens and workers were dissected under a binocular microscope (SZ40; OLYM-PUS) to determine the number of ovarioles. All statistical tests were conducted using R software (ver. 3.3.1; R Development Core Team 2013). Voucher specimens were deposited in the Labo-ratory of Insect Ecology at Kyoto University and the Ehime University Museum.

RESULTS

Colony composition: Twenty-three colonies of L. daikoku were collected between 5 May and 30 September (Table 1). The mean colony size was 12.74 ± 12.60 workers (mean ± SD, n = 23), ranging from 1 (a founding queen) to 44. Of the 23 colonies, 19 (82.6%) had a single queen (monogynous), 3 (13.0%) had multiple queens (polygynous), and 1 (4.3%) had no queen (Table 1). Immatures at different stages coexisted in the colonies. Alates of both sexes were first ob-served in the two colonies collected on 16 May, and sexuals were continuously collected until 30 September when the last L. daikoku colonies were collected. The largest number of sexuals (48) was found in colony 6 (Table 1). The popu-lation size of colonies producing sexuals ranged from 5 to 44 workers. No relationship was found between colony size and the number of sexuals produced (Poisson regression, likelihood ratio test, χ2 = 0.279, df = 1, p = 0.598, n = 9; Fig. 1). Most of these colonies produced a large num-ber of sexuals relative to their colony size (Fig. 1). There were two colonies that had a single gyne (10 & 14) and without these two colonies, a significantly positive relationship was detected between colony size and the number of sexuals produced (Poisson regression, likelihood ratio test, χ2 = 23.985, df = 1, p < 0.001, n = 7). We could not determine whether these two colonies were irregular. Further collections are required to determine if these colonies are highly unusual or not. The proportion of males to the total number of sexuals ranged from 0 to 0.65, with a mean value of 0.30. Solitary queens (presumably at the founding stage as these colonies did not contain any brood) were found on 27 May, 3 June, and 2 July. These queens also nested in cavities in bam-boo stems.

Caste dimorphism: Significant size dimorphism was observed between queens and workers in the following body regions: head width, 0.50 ± 0.02 mm for queens (mean ± SD, n = 43, range 0.45–0.56) and 0.48 ± 0.02 mm for workers (mean ± SD, n = 48, range 0.44–0.51; Welch’s t-test, t = −5.38, df = 69.41, p = 0.001); thorax width, 0.45 ± 0.03 mm for queens (mean ± SD, n = 43, range 0.40–0.53), and 0.37 ± 0.02 mm for workers

3Life history of a basal doryline ant

(mean ± SD, n = 48, range 0.34–0.41; Welch’s t-test, t = −15.77, df = 63.40, p = 0.001); and ab-domen width, 0.57 ± 0.03 mm for queens (mean ± SD, n = 43, range 0.52–0.66) and 0.52 ± 0.02 for workers (mean ± SD, n = 48, range 0.46–0.56; Welch’s t-test, t = −8.95, df = 68.36, p = 0.001). In addition, there was a marked queen-worker difference in the adult thorax structure (Fig. 2A & B): the queen’s thorax had clear boundaries (i.e., sutures) between the pronotum, mesonotum, and metanotum (Fig. 3A & C), and normal wings were initially attached to the latter two segments. The worker’s dorsal thoracic sclerites were fused completely (Fig. 3B & D). Female castes also dif-fered in the number of ovarioles: queens (n = 43) had four (2 + 2), whereas all but one (n = 48) worker had two (1 + 1); one “worker”, as deter-mined by external morphology, had four ovari-oles (2 + 2).

Table 1. Composition of the Lioponera daikoku colonies and Monomorium brood found in the nests.

L. daikoku Monomorium spp.Colony

ID Date Queen Worker Male Gyne Cocoon Larva Egg Larva Pupa

1 May 5 1 2 - - - - - - -2 May 11 1 13 - - - - - - -3 May 11 1 1 - - - - - - -4 May 16 1 15 - - - - - - -5 May 16 1 34 30 16 - - - 60 -6 May 16 1 20 16 32 - 1 1 202 -7 May 20 1 29 - - - 30 4 70 -8 May 27 1 - - - - - - - -9 May 27 1 - - - - - - 20 -10 May 27 1 44 - 1 - 29 - 150 -11 June 3 1 - - - - - - - -12 July 2 1 - - - - - - 10 113 July 2 3 14 - - - 6 7 102 4014 July 2 2 24 - 1 14 36 29 42 -15 July 2 1 20 - - 3 27 13 112 3416 July 25 5 - - - 33 39 35 105 517 July 25 1 3 - - 10 4 6 59 1418 Aug. 8 1 18 - 23 47 23 10 177 -19 Aug. 19 - 5 2 3 4 - - 1 -20 Sep. 19 1 27 14 31 - 58 5 123 -21 Sep. 30 1 9 14 17 - 14 - 53 -22 Sep. 30 1 3 - - 4 12 2 7 1423 Sep. 30 1 12 20 15 - 47 4 108 4

Fig. 1. Relationship between colony size and the num-ber of sexuals. Most colonies had a larger number of sexuals compared to the number of workers (i.e., above the dashed line).

4 Naoto Idogawa & Shigeto Dobata

Fig. 2. Dorsal view of L. daikoku female castes: A gyne, B worker. Scale bar: 1.0 mm.

Nesting and food habits: Six ant species were found inside bamboo cavities at the study site: Monomorium triviale Wheeler, 1906 (32 nests), Camponotus yamaokai Terayama & Satoh, 1990 (22 nests), Temnothorax spp. (20 nests), Crema-togaster teranishii Santschi, 1930 (9 nests), Doli-choderus sibiricus Emery, 1889 (3 nests) and M. intrudens Smith, 1874 (1 nest), together with six L. daikoku colonies. The first three species occu-pied the majority of nests (74/87 = 85%). The di-ameters of nest stems of Ca. yamaokai were sig-nificantly larger (Tukey–Kramer test, p ≤ 0.0046) than for the other species (including L. daikoku), which showed no significant difference among them (p ≥ 0.252; Fig. 4).

Throughout the field study, larvae and pupae of unidentified species of Monomorium were found in the nests of L. daikoku (Table 1 and Fig. 5). These immatures likely belonged to M. intrudens and M. triviale because these Monomorium occur at the study site. Feeding experiments conducted in the laboratory revealed that the workers of L. daikoku stung the larvae of M. intrudens and M. triviale and carried them back to their nests (Fig. 6), while they did not do this when provided with larvae of Ca. yamaokai and Temnothorax spp.

Male morphology: Twelve specimens were ex-amined to obtain the following measurements: Total body length, 2.94 ± 0.17 mm (range: 2.63–

5Life history of a basal doryline ant

Fig. 3. Thorax structures of L. daikoku female castes: A dorsal view of gyne (wings removed to show thorax), B dorsal view of worker, C lateral view of gyne, D lateral view of worker. Scale bar: 0.5 mm.

Fig. 4. Box plot illustrating the diameters of bamboo stems containing nests of the seven ant species. Each box shows upper and lower quartiles along with maximum and minimum (whiskers), and median (thick line). The numbers in parentheses under the species name indicate the total numbers of measured nest stem. Different letters indicate significant differences (Tukey–Kramer test, p < 0.05).

6 Naoto Idogawa & Shigeto Dobata

Fig. 5. Larvae (white arrowheads) of Monomorium found among a L. daikoku larvae (black arrowheads) inside a field collected nest. Because they disappeared in a few weeks, it was suspected that they were consumed by adults or larvae of L. daikoku.

Fig. 6. Foraging behavior of L. daikoku workers in the laboratory. They bit and stung the Monomorium triviale larvae that were fed experimentally.

7Life history of a basal doryline ant

3.26). Head (Fig. 7A): head width across eyes, 5.01 ± 0.01 mm (range: 0.49–0.53); head length, 0.56 ± 0.02mm (range: 0.54–0.58); eye width, 0.19 ± 0.01 mm (range: 0.16–0.22); eye length, 0.25 ± 0.02 mm (range: 0.21–0.29); and scape length, 0.20 ± 0.01 mm (range: 0.18–0.21). Me-sosoma (alitrunk) and abdomen (Fig. 7B and C): mesosoma width in dorsal view, 0.55 ± 0.02 mm (range: 0.52–0.58); mesosoma length, 1.04 ± 0.05 mm (range: 0.94–1.10); mesosoma height, 0.47 ± 0.02 mm (range: 0.44–0.51); petiole width in dorsal view, 0.32 ± 0.02 mm (range: 0.28–0.35); and petiole length, 0.27 ± 0.03 mm (range: 0.23–0.33); petiole height, 0.28 ± 0.02 mm (range: 0.26–0.34). Right wings (Fig. 7D and E): fore-wing width, 0.67 ± 0.02 mm (range: 0.63–0.69); forewing length, 2.23 ± 0.05 mm (range: 2.13–2.35); hindwing width, 0.40 ± 0.01 mm (range:

0.38–0.42); and hindwing length, 1.75 ± 0.06 mm (range: 1.60–1.83). The males of Lioponera are character-ized by (1) antennae with 13 segments, (2) the absence of a costal vein on the forewing, and (3) the presence of a free stigmal vein on the fore-wing (Borowiec, 2016). All of these characteris-tics were shared with the males of L. daikoku. As in other Lioponera species, L. daikoku also had edentate mandibles, a single spur on the middle tibia, and a large brown pterostigma on the fore-wing. Most of the body was black, while the man-dibles, antennae, and legs were yellowish brown. The veins on the wings were very pale. Although the male of L. daikoku resembles that of L. longi-tarsus, the type species of this genus, the former is characterized by a V-shaped groove (=notau-lus) on the mesoscutum, and black body color.

Fig. 7. External morphology of a male L. daikoku: A head in full-face view, B lateral view, C dorsal view, D right forewing, E right hindwing. Scale bar: 1.0 mm.

8 Naoto Idogawa & Shigeto Dobata

DISCUSSION

The colony size (number of workers) of Dor-ylinae colonies varies markedly among species, ranging from 9.1 in Syscia humicola (Masuko 2006) and 150–600 in Ooceraea biroi (Tsuji and Yamauchi 1995), both belonging to the former Cerapachyinae, to > 20 million in some Dorylus army ants (e.g., Vosseler 1905, Raignier & Boven 1955). The small colony size of L. daikoku of about 10 on average is the smallest among previ-ous single-nest records of colony size of Liopo-nera species (ranging from a few dozen to sev-eral hundred, Clark 1924; Wilson 1958), and is of interest regarding the basal status of this species in the Dolyrinae. Monogyny was the dominant social structure of L. daikoku. However, three of 23 colonies had more than one dealate queen, whose reproductive status was unknown. One of them, collected on 25 July 2017 (Colony 16), had five dealate queens and many brood, despite the lack of workers (Table 1). Although these queens were not dissected, this hints at pleometrotic colony foundation. Clark (1924) and Hölldobler (1982) reported polygynous nests in other species of Lioponera, suggesting the existence of within- and between-species diversity of social structure. Synchronization of brood production was not ob-served in L. daikoku, indicating that this species is non-phasic (see also Ito et al. 2018). Gotwald (1988) postulated that the phasic lifestyle of army ants and clonal raider ants evolved secondarily from non-phasic ancestors. The updated phyloge-ny (Borowiec 2017) and studies of the life cycles of Lioponera (reviewed in Borowiec 2016) sup-port this hypothesis. Our study suggests that the mating flight occurs in May, and involves alate sexuals that emerged the previous August. Soli-tary dealate queens were found in the field, sug-gesting that L. daikoku employs independent col-ony founding. As some Lioponera species have ergatoid queens and reproduce by colony fission (Borowiec 2016), and a comparative study of the mode of colony founding in this genus would be of interest. Our field and laboratory observations strongly suggest that L. daikoku is a special-ist predator of ant brood, especially of Mono-morium. Although Lioponera are specialist ant hunters, the developmental stage of their prey

varies among species. Hölldobler (1982) and Brown (1975) observed Lioponera workers prey-ing on both adult and immature Pheidole sp. We observed in the laboratory that adult work-ers of Monomorium were stung and killed by L. daikoku but were not consumed, suggesting that L. daikoku preys exclusively on brood. At the study site, Monomorium is the only genus with an adult body size smaller than L. daikoku (data not shown), which might explain the preference for Monomorium. There were two Monomorium species at the study site, and we postulate that M. triviale is the dominant prey of L. daikoku due to its relative abundance and smaller colony size. We did not observe the foraging behavior of L. daikoku in the field. T. Satoh (pers. comm.) ob-served several L. daikoku workers in a queue on a tree branch, seemingly engaged in group for-aging. Some studies of other Lioponera species reported group foraging (reviewed in Borowiec 2016). Interestingly, the number of fresh brood items in the nests of L. daikoku seemed to far ex-ceed the capacity of the workers to carry them (Table 1). In Thailand, R .Mizuno (pers. comm.) observed that a colony of Cerapachys sp. moved into the abandoned nests of prey ant species, possibly to predate on the remaining brood. The foraging behaviors of L. daikoku and other basal doryline species deserve further study. Finally, we found a low degree of di-morphism between queens and workers, and monomorphism among workers in L. daikoku, which is also the case in the basal Dorylinae groups. A female with the external morphology of a worker and the same number of ovarioles as a queen can be classified as an intercaste (Peeters 1991, Molet et al. 2012); this has been reported in Lioponera clarki Crawley, 1922 (Clark 1924) and might reflect the low degree of queen–worker dimorphism.

ACKNOWLEDGEMENTS

We thank Dr. M. Terayama (University of Tokyo), Dr. T. Satoh (Tokyo University of Agriculture and Technology), Mr. Yu Hisasue (Ehime University) and Dr. M. Yoshimura (Okinawa Institute of Sci-ence and Technology Graduate University) for their valuable comments and Dr. K. Matsuura

9Life history of a basal doryline ant

(Kyoto University) for laboratory support. Dr. F. Ito and Mr. R. Mizuno (Kagawa University) pro-vided helpful information on the ecology of Lio-ponera ants. Dr. A. Cronin (Tokyo Metropolitan University), Dr. C. Peeters (IEES-Paris) and an anonymous referee are thanked for helpful contri-butions toward improving the manuscript.

REFERENCES

antmaps.org, 2018. Dorylinae. Available from http://antmaps.org/?mode=diversity&subfamily=Dorylinae (accessed on 16 Jan. 2018).

Borowiec ML, 2016. Generic revision of the ant sub-family Dorylinae (Hymenoptera, Formici-dae). ZooKeys 2016(608): 1–280.

Borowiec ML, 2017. Convergent evolution of the army ant syndrome and congruence in big-data phylogenetics. bioRxiv: 1–61.

Brady SG, Fisher BL, Schultz TR, & Ward PS, 2014. The rise of army ants and their relatives: di-versification of specialized predatory doryline ants. BMC Evolutionary Biology 14(1): 93.

Brown WL, 1975. Contributions towards a reclassifi-cation of the Formicidae. V. Ponerinae, tribes Platythyreini, Cerapachyini, Cylindromyr-mecini, Acanthostichini, and Aenictogitini. Search Agriculture (Ithaca) 5: 1–115.

Camazine S, Deneubourg J-L, Franks N, Sneyd J, Bo-nabeau E, Theraulaz G. 2001. Self-Organi-zation in Biological Systems. Princeton Uni-versity Press.

Clark J 1924, Australian Formicidae. Journal and Proceedings of the Royal Society of Western Australia 9: 72–89.

Gotwald WHJ, 1988. On becoming an army ant. In: Advances in Myrmecology (Trager JC, ed.), Brill, Leiden, 355–370.

Gotwald WHJ, 1995. Army Ants: The Biology of So-cial Predation. Cornell University Press, Ithaca, 302 pp.

Hölldobler B, 1982. Communication, raiding behavior and prey storage in Cerapachys (Hymenop-tera: Formicidae). Psyche (Cambridge) 89: 3–23.

Ito F, Jaitrong W, Hashim R, Mizuno R, 2018. Colony composition, brood production and caste di-morphism in two species of the Oriental dory-line, Lioponera suscitata group (Formicidae: Dorylinae). Asian Myrmecology xx: xx-xx.

Kronauer DJC, 2009. Recent advances in army ant bi-ology (Hymenoptera: Formicidae). Myrme-cological News 12: 51–65.

Masuko K, 2006. Collection and the result of dissection of the ant Cerapachys humicola. Ari 28: 1–6.

Molet M, Wheeler DE, Peeters C, 2012. Evolution of novel mosaic castes in ants: modularity, phenotypic plasticity, and colonial buffering. The American Naturalist 180 (3): 328-341.

Mayr G, 1879. Beiträge zur Ameisen-Fauna Asiens. Verhandlungen der Kaiserlich-Königlichen Zoologisch-Botanischen Gesellschaft in Wien 28:645-686.

Oxley PR, Ji L, Fetter-Pruneda I, McKenzie SK, Li C, Hu H, Zhang G, Kronauer DJ, 2014. The genome of the clonal raider ant Cerapachys biroi. Current Biology 24 (4): 451–8.

Peeters CP, 1991. Ergatoid queens and intercastes in ants: two distinct adult forms which look morphologically intermediate between workers and winged queens. Insectes So-ciaux 38(1): 1–115.

R Core Team 2018. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL http://www.R-project.org/ (accessed on 16 Jan. 2018)

Raignier A, Boven JKA 1955 Étude taxonomique, biologique et biométrique des Dorylus du sous-genre Anomma (Hymenoptera Formi-cidae). Annales du Musée Royal du Congo Belge. Nouvelle Série in Quarto. Sciences Zoologiques 2: 1–359.

Schneirla TC, 1971. Army Ants: A Study in Social Or-ganization. W H Freeman and Company, San Francisco 349 pp.

Terayama M, 1996. Taxonomic Studies on the Japa-nese Formicidae, Part 2. Nature and Human Activities 1: 9–32.

Terayama M, Kubota S, Eguchi K, 2014. Encyclopedia of Japanese Ants. Asakura Shoten, Tokyo, 278pp (In Japanese).

Tsuji K and Yamauchi K, 1995. Production of females by parthenogenesis in the ant Cerapachys bi-roi. Insectes Sociaux 42: 333–336.

Vosseler J, 1905. Die ostafrikanische Treiberameise (Siafu). Pflanzer 1: 289–302.

Wilson EO, 1958. Observations on the behavior of the cerapachyine ants. Insectes Sociaux 5(1): 129–140.

ASIAN MYRMECOLOGYA Journal of the International Network for the Study of Asian Ants

Communicating Editor: Adam Cronin